Bordetella pertussis is an important pathogen of children and increasingly also of adults worldwide (1,2). For unknown reasons, whooping cough does not respond to antibiotics and the substantial beneficial health impact of the diphtheria-pertussis-tetanus vaccine is achieved at the cost of the toxicity of the controversial pertussis component. It has thus become important to investigate alternative modes of intervention in disease which could have improved acceptability without sacrificing efficacy. As many steps in the pathogenesis of whooping cough are still undefined, rational design of vaccine improvements is difficult. However, since the disease in entirely noninvasive, an understanding of the biology of the interactions between the bacterial and host cell surfaces is very likely to be of major importance to efforts to interrupt disease. This proposal addresses the mechanisms of two such interactions: the specific adherence of virulent B. pertussis to human cilia and bacterial lysis by cell wall-active antibiotics. A significant contribution to the understanding of the mechanism of adherence of B. pertussis to cilia resulted from the development of a unique in vitro adherence assay using human ciliated cells (the appropriate target for an exclusively human infection) (3). Two antigenic bacterial surface proteins have been found to contribute to adherence (4). This proposal seeks to determine the structure of the carbohydrate-containing human ciliary receptor for the bacterial adhesins and define an analog of the receptor which may act as an effective competitive inhibitor of adherence in vivo. Assessment of the protective efficacy of adhesin and receptor components has the potential to define antiadherence activities important to new prophylactic and therapeutic agents for whooping cough. In light of the suggestion that a cell wall fragment of B. pertussis comprises the tracheal cytotoxin which specifically kills ciliated cells (5), it is now important to define how cell wall-active antibiotics interact with this organism. In this proposal, the penicillin binding proteins will be characterized in terms of number and relative binding affinity to selected beta lactam antibiotics. The slow bactericidal response of B. pertussis to antibiotics will be assessed in view of its importance as an example of phenotypic tolerance of all slowly growing bacteria. Antibiotics with unusually high bactericidal activity against slowly growing bacteria (6) will be tested against B. pertussis. Identification of antibiotics capable of rapidly killing but not lysing B. pertussis (i.e., no release of toxic cell wall) may provide a basis for improved therapy of whooping cough.